Fig 2: RNA expression profiling identifies MIST1 as a DDX56-downstream target in the regulation of the PTEN-AKT axis. (A) A volcano plot constructed using P-values and fold-change to illustrate the differential downstream target genes between DDX56-silenced MHCC97H cells and control cells (|fold change| =2, P < 0.05). Red dots represent significantly upregulated genes while blue dots represent significantly downregulated genes. (B) The top 20 differential downstream genes from RNA-seq results were validated by RT-PCR in MHCC97H cells. (C-D) The mRNA and protein levels of MIST1 in both MHCC97H and Huh-7 cells were detected using RT-PCR and western blotting, respectively. (E) IHC staining shows the expression level of MIST1 in the subcutaneous tumor. Scale bars: 250 µm (top), 50 µm (bottom). (F) Representative images of DDX56 and MIST1 immunohistochemical staining in human HCC tissues (left). Correlation analysis between DDX56 and MIST1 protein expression in HCC tissues (right). Scale bars: 250 µm (low magnification), 50 µm (high magnification). (G) Effect of MIST1 knockdown on MHCC97H and Huh-7 cell proliferation was determined by colony formation assay. (H) The indicated protein was determined using a western blot analysis of MHCC97H and Huh-7 cells infected with MIST1-silencing lentivirus (shMIST1) or lentivirus control (shCont). (I) Western blot analysis of the PTEN-AKT axis in MIST1-overexpressed HCC cells, showing that PTEN overexpression reversed the effect of MIST1 on the PTEN-AKT axis. Cells were transfected with 3 µg of the indicated plasmids for 72 h.

Fig 3: Effect of DDX56 in promoting HCC cell proliferation in vitro. (A) The mRNA and protein level of DDX56 in HCC cells and immortalized normal liver cells (L02) as determined using RT-PCR and western blot analyses, respectively. (B) The lentiviral-mediated shRNA silencing effects of DDX56 in MHCC97H and Huh-7 cells were examined using western blotting. (C-D) The effect of DDX56 knockdown on HCC cell proliferation was determined by CCK8 (C) and colony formation assays (D). (E) MHCC97H and Huh-7 cells were seeded onto coverslips and DNA synthesis was assessed via EdU immunofluorescence staining. The graph on the right depicts the percentage of EdU-positive nuclei. Scale bars: 100 µm. (F-H) Effect of DDX56 gain-of-function on PLC/RPF/5 and Hep3B cell proliferation was determined by CCK8 (F), colony formation assay (G), and EdU assay (H). Scale bars: 100 µm.

Fig 4: DDX56 promotes MIST1 transcription by recruiting the epigenetic regulator MECOM. (A) Co-immunoprecipitation experiment showed the interaction between DDX56 and MECOM in MHCC97H and Huh-7 cells. (B) Immunofluorescence analysis shows the co-localization of DDX56 (green) and MECOM (red). (C) DDX56 rescues MECOM to regulate MIST1 promoter-driven luciferase activity. DDX56-knockdown HCC cells were co-transfected with MIST1 promoter luciferase constructs and MECOM overexpression plasmid. Cell lysates were measured for firefly and Renilla (RL-TK) activity. (D-E) ChIP-qPCR assay with anti-MECOM or anti-H3K9me1 was performed in MHCC97H and Huh-7 cells. DDX56 loss-of-function reduced MECOM (D) or H3K9me1 (E) occupancy at the promoter of MIST1 gene (top). Agarose gel electrophoresis of PCR fragments after ChIP (bottom). (F) Western blot analysis of MIST1/PTEN/AKT axis in DDX56-silenced HCC cells, and MECOM overexpression reversed the effect of DDX56 on the MIST1/PTEN/AKT axis. (G) EdU assays revealed that ectopic MECOM expression in DDX56-silenced cells partially reversed the DDX56 knockdown-mediated inhibitory effect on HCC proliferation. Scale bars: 100 µm.

Fig 5: DDX56 is overexpressed in patients with HCC and correlates with the poor prognosis subtype. Venn diagrams depicting (A) 1306 commonly upregulated genes (Up) and 349 downregulated genes (Down) based on RNA-seq data from TCGA-LIHC and four paired RNA-seq data from our laboratory (|fold change| = 2, P < 0.05); (B) 132 commonly upregulated and 6 downregulated RNA-binding protein (RBP) genes; (C) 9 candidates that were identified based on an analysis of the top 200 survival-associated genes, tumor grade-associated genes, and upregulated RBPs (P < 0.05). (D) Heatmap showing the expression profiles of these 9 candidate RBPs in four liver cancer tissues and paired adjacent tissues (|fold change| = 2, P < 0.05). (E-F) Relative mRNA levels of DDX56 in GSE14520 and GSE25097 (***, P < 0.0001) (E) and in the TCGA-LIHC database (***, P < 0.0001) (F). (G) The mRNA level of DDX56 was significantly associated with multiple clinicopathologic parameters, including AFP, T classification, and histologic grade (*, P < 0.05, **, P < 0.01). (H) The DDX56 high expression group showed worse overall survival (OS) and disease-free survival (DFS) in liver cancer patients (OS, P = 0.0004; DFS, P = 0.00013, Log-rank test). (I) Representative western blot images of DDX56 in HCC tissues and the corresponding adjacent normal tissues. (J) Representative immunohistochemical staining images of DDX56 from the HCC tissue array obtained from Servicebio (n = 32). Scale bars: 500 µm (top), 50 µm (bottom).